Combinatorics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
Wednesday 22 April
Rex, short for Reverse Hex, is a set coloring game in which players try to avoid connecting terminals of their color. Combinatorial game theory (CGT) is the study of perfect strategy games. Until recently, both Rex and Hex were not examined through the lens of CGT. In this thesis we take inspiration from the study of normal play games by Berlekamp, Conway, and Guy, along with the combinatorial game theory of Hex developed by Selinger, to develop methods for analyzing Rex positions. We explore how to tell if one position is preferable to another, how to simplify positions, and some special properties of Rex (and antimonotone set coloring games in general). By the end of this thesis we will be able to take a position in a game of Rex, break it into smaller positions, analyze each of the smaller positions, then add the results back together to more easily determine who wins and loses the larger position.
Classical Analysis and ODEs
A note on the sum-product problem for fractal sets
Utilising recent advances in incidence geometry for balls and tubes, and advances in sum-product theory in the discrete setting, we show that for $0 < s \leq 1/2$ and for any $A \subset \mathbb{R}$ with Hausdorff dimension $s$, either the upper-box dimension of $AA$, or the lower-box dimension of $A+A$ must be at least $29s/23$. We obtain the slightly better bound of $33 s / 26$ when we replace the sum-set with the smoother difference-set.
Read the full abstract
Group Theory
Numerical Semigroups with $a_e = 2g+1$
This article discusses numerical semigroups having a generator which is as large as possible. This turns out to be $2g+1$, where $g$ is the genus of the semigroup. We will show that these semigroups are closely related to symmetric semigroups and have interesting symmetry properties themselves. Furthermore we will show that Wilf's question has a positive answer for these semigroups and some semigroups derived thereof.
Read the full abstract
General Mathematics
Introduction to generalised Cesaro convergence II
In this second of three introductory papers, we extend the notion of generalised Cesaro summation/convergence to the more natural setting of what we call remainder Cesaro summation/convergence. This greatly expands the range of problems susceptible to Cesaro methods and introduces the geometric location of summands as a critical consideration. We also show that geometric generalised Cesaro convergence is invariant under dilation and scaling. We present a number of calculations illustrating the utility of these developments. In particular we introduce a new, more natural definition of the classical Gamma function using remainder Cesaro summation/products, and show that many its key properties - both basic and advanced - fall out directly and intuitively from this Cesaro definition and its geometric and dilation-invariance properties. We also consider other examples and show how Cesaro methodology explains the common structure of many well-known functional equations.
Read the full abstract
General Mathematics
An Elementary Proof Of The Chebyshev Bounds Via A Weighted Prime Sum
The Chebyshev bounds for the prime-counting function, i.e., $\pi(x) \asymp x/\ln x$, is established in a new way. This new approach follows the outline $\pi(x) \asymp \bigl(\sum_{p \le x} \sqrt{\ln p / p}\bigr)^{2} \asymp x/\ln x$. Here, the second $\asymp$ is derived from the classical estimate by Mertens, i.e., $\sum_{p \le x} (\ln p)/p = \ln x + O(1)$; while the first $\asymp$ is proved by considering the difference $\bigl(\sum_{p \le x} \sqrt{\ln p/p}\bigr)^{2} - \sum_{p \le x} (\ln p)/p$, which is shown as having the same order as $\pi(x)$.
Read the full abstract
Number Theory
Relations between higher level Hurwitz class numbers
We connect generalizations of the classical Hurwitz class numbers coming from two different frameworks: one introduced by Pei and Wang, arising from the generalized Cohen–Eisenstein series, and another by Li, Skoruppa, and Zhou, arising from Eichler orders of quaternion algebras. As applications, we obtain new basis for Eisenstein space $E_{3/2}^{+}(4N,\mathrm{id})$, a generalization of recent results of Beckwith and Mono, and a generalization of Gauss' formula.
Read the full abstract
Combinatorics
Intersection numbers of Sudoku latin squares
Let $n=hw$, where $h$ and $w$ are integers with $h,w \ge 2$. We determine the set of possible intersection numbers of two $n \times n$ latin squares having the additional `Sudoku' constraint based on a $w \times h$ grid of $h \times w$ boxes.
Analysis of PDEs
Liouville Type Results for Quasilinear Elliptic Inequalities Involving Gradient Terms on Weighted Graphs
In this paper, we study the following quasi-linear elliptic inequality $\Delta_m u +u^p |\nabla u|^q \leqslant 0$ on weighted graphs, where $(m,p,q)\in (1,\infty)\times\mathbb{R}\times\mathbb{R}$. According to the ranges of parameters $(m, p, q)$, we establish the non-existence of nontrivial positive solutions under the corresponding sharp volume growth conditions. Our results can be viewed as a discrete generalization of their counterparts on Riemannian manifolds established by [Sun, Yuhua; Xiao, Jie; Xu, Fanheng, Math. Ann. 384 (2022), no. 3-4, 1309–1341.]. However, this generalization is far from trivial, many results exhibit significant differences from the manifold setting, highlighting the distinct behaviors and challenges that arise in the discrete weighted graph framework.
Read the full abstract
Probability
Quantitative homogenization for the critical long-range random conductance model
We consider the long-range random conductance model on $\mathbb{Z}^d$ at the critical exponent: the jump rate between sites $x$ and $y$ decays as $\mathbf{a}(x,y) |x-y|^{-(d+2)}$, where $\mathbf{a}(x,y)$ are i.i.d. uniformly elliptic conductances. Below the critical exponent $(d+2)$ the walk converges to a stable process; above it, to Brownian motion with diffusive $\sqrt{t}$ scaling. At criticality the second moment of the jump kernel diverges logarithmically. We establish quantitative homogenization of the associated elliptic equation to the Laplacian at the rate $1/\sqrt{|\ln\varepsilon|}$. As a consequence, we deduce quenched convergence of the random walk to Brownian motion under the anomalous $\sqrt{t \log t}$ scaling. Unlike in standard homogenization, the effective diffusivity is determined by the mean conductance alone, with no corrector contribution at leading order.
Read the full abstract
Probability
Gaussian fluctuations for Internal DLA on cylinders
Internal DLA is a discrete random growth model describing growing clusters of particles. Its limiting shape and fluctuations are well understood when the underlying graph is the $d$-dimensional lattice or the cylinder $\mathbb{Z}_N \times \mathbb{Z}$. In the latter geometry, the average fluctuations of IDLA have been shown to converge to the GFF. In this note we generalise this result by showing that, for any vertex-transitive base graph $V_N$ satisfying an eigenvalue convergence condition, the average fluctuations of IDLA on the cylinder $V_N \times \mathbb{Z}$ are given by a GFF. On the way, we present an improved bound on the clusters' maximal fluctuations, which is of independent interest and which implies a shape theorem for IDLA on $V_N \times \mathbb{Z}$ for any vertex-transitive base graph $V_N$.
Read the full abstract
Functional Analysis
On measurability of Kurzweil–Stieltjes integrable functions on compact lines
We continue the study on Kurzweil–Stieltjes integration on compact lines initiated in [doi:10.1007/s11117-025-01161-9]. Given a real valued function $G$ on a compact line, the presented integral is called the Kurzweil–Stieltjes integral with respect to $G$, or simply the $G$-integral. %Given a compact line $K$ and a right-continuous function $G:K\to\mathbb{R}$ of bounded variation, we consider the Radon measure $\mu_G$ naturally induced by $G$. Our main results concern the relationship between $G$-integrability and measurability. We prove that, whenever $G$ is nondecreasing, every $G$-integrable function is $\mu_G$-measurable, where $\mu_G$ is the natural Radon measure induced by $G$. We also show that, for an arbitrary $G$ of bounded variation, every bounded $G$-integrable function is $\mu_G$-measurable. %, where $|\mu_G|$ denotes the total variation measure of $\mu_G$. As an application, we provide a full characterization of Lebesgue integrablility with respect to Radon measures in terms of the $G$-integral, and demonstrate that the $G$-integral represents an extension of the Lebesgue integral with respect to $\mu_G$ for suitable $G$. In addition, we establish a version of Hake's theorem for the $G$-integral in this setting.
Read the full abstract
Combinatorics
Some Results on Additively Graceful Signed Stars and Double Stars
We study additively graceful labelings of signed graphs on stars and double stars. While the case of signed stars is straightforward, the problem becomes significantly more intricate for signed double stars. We obtain a characterization of additively graceful signed stars, while, for several sub-classes of additively graceful signed double stars, we establish existence, uniqueness and non-existence results.
Read the full abstract
Optimization and Control
The Method of Ellipcenters for Strongly Convex Functions
The Method of Ellipcenters (ME), introduced in~\cite{ME2025} for strongly convex quadratic minimization, uses two gradient evaluations per iteration: one at the current iterate and one at a companion point on the same level set. We extend ME to the broader class of strongly convex functions with Lipschitz continuous gradient. We prove that ME contracts unconditionally at the linear rate $1-\mu^2/L^2$, and that at every step where the two gradient directions are linearly independent, which, in dimension at least two, is every step generically, it matches the rate of gradient descent with exact line search. In that linearly independent case, a midpoint argument exploiting the level-set symmetry yields a further per-step improvement, which is global when the angle between the two gradients is uniformly bounded away from zero. The same symmetry forces this angle to be obtuse, so the improvement is strictly active at every such step. ME also converges in at most two steps in dimension two. Numerical experiments on regularized logistic regression confirm the theoretical predictions.
Read the full abstract
Number Theory
Extremal Problems for GCDs and LCMs in Higher Dimensions
We study extremal problems for tuples of integers chosen from sets $A_i \subset [X_i,2X_i]$ for $1\le i\le k$, under large GCD and small LCM conditions. For the GCD problem, we extend the work of Green and Walker to higher dimensions. Specifically, for $k\ge 3$, if $\gcd(a_1,\dots,a_k)\ge D$ for at least a proportion $\delta$ of the tuples in $\prod_{i=1}^k A_i$, then $$ \prod_{i=1}^k |A_i| \ll_{k,\varepsilon} \delta^{-k/(k-1)-\varepsilon} \frac{\prod_{i=1}^k X_i}{D^k}. $$ The proof is based on a minimal counterexample argument and a new high-dimensional measure concentration lemma. We also establish a large sieve-type inequality to obtain a complementary estimate for the GCD problem. For the LCM problem, we use a quite different method to show that, for all $k\ge 2$, $$ \prod_{i=1}^k |A_i| \ll_{k,\varepsilon} \delta^{-k/(k-1)} \frac{L^{k/(k-1)+\varepsilon}} {\bigl(\prod_{i=1}^k X_i\bigr)^{1/(k-1)}}, $$ whenever $\operatorname{lcm}(a_1,\dots,a_k)\le L$ for at least a proportion $\delta$ of the $k$-tuples in $\prod_{i=1}^k A_i$. Finally, we show that these bounds are essentially best possible up to $\varepsilon$-losses in the exponent.
Read the full abstract
Operator Algebras
Uniqueness theorems for combinatorial C*-algebras
Spielberg's construction of C*-algebras from left cancellative small categories is a common generalization for most C*-algebras one would consider to come from “combinatorial data,” including graph and $k$-graph C*-algebras, Li's semigroup C*-algebras, Nekrashevych's self-similar action algebras, and more. We use known groupoid models of these algebras and Exel's theory of tight representations of inverse semigroups to prove uniqueness theorems for these C*-algebras. As applications, we improve on our previous uniqueness theorem for the boundary quotient C*-algebras of right LCM monoids, and we also generalize the uniqueness theorem of Brown, Nagy, and Reznikoff for row-finite higher-rank graphs to the finitely aligned case.
Read the full abstract
Differential Geometry
Special Lagrangians with Cylindrical Tangent Cones
We construct new examples of special Lagrangian submanifolds $Y\subset \mathbf{C}^{n+1}$, $n\geq 3$ in a neighborhood of the origin, with an isolated singularity, but with cylindrical tangent cone $C\times\mathbf{R}$. Moreover, $Y\setminus\{0\}$ is connected while $(C\setminus\{0\})\times\mathbf{R}$ is not. Such examples exist, for example, when $C$ is a pair of transverse planes.
Read the full abstract
Probability
Tail Bounds via Southwest Boundary
We derive upper bounds for probabilities of the form $P(g(\mathbf{X})\geq t)$ using the southwest boundary (recently introduced in our previous work) $\partial_{\mathrm{SW}} Q(g^{-1}[t,\infty))$, where $Q$ is a reflection to the first quadrant. Under natural continuity, symmetry, and monotonicity assumptions on $g$, this yields explicit and computable bounds of the form $P(g(\mathbf{X})\ge t)\le ns_t$, where $s_t$ is the unique parameter at which the line $L(s)=(f_1^{-1}(s),\dots,f_n^{-1}(s))$ intersects the southwest boundary. In particular, when $g$ is a homogeneous polynomial of degree $k$ (plus a constant $C$) and all tail bounds on the random variables are identical, the bound proves to the closed-form expression $$ P(g(\mathbf{X})\ge t)\leq nf\bigg(\frac{(t-C)^{1/k}}{(\sum_i|a_i|)^{1/k}}\bigg) $$ where $a_i$ are the coefficients of the monomials in $g$. We then obtain an explicit tail bound for the trace of a Schur multiplier acting on random matrices with identical tail bounds on the random variables. No assumptions are made about independence or dependence.
Read the full abstract
Statistics Theory · Cross-listed from stat.ME
A goodness-of-fit test for the logistic propensity score model under nonignorable missing data
Logistic regression is widely used to model the propensity score in the analysis of nonignorable missing data. However, goodness-of-fit testing for this propensity score model has received limited attention in the literature. In this paper, we propose a new goodness-of-fit testing procedure for the logistic propensity score model under nonignorable missing data. The proposed test is based on an unweighted sum-of-squared residuals constructed from the marginal missingness mechanism and accommodates the partial observability of the outcome. We establish the asymptotic distribution of the test statistic under both the null hypothesis and general alternatives, and develop a bootstrap procedure with theoretical guarantees to approximate its null distribution. We show that the resulting bootstrap test attains asymptotically correct size and is consistent, with power converging to one under model misspecification. Simulation studies and a real data application demonstrate that the proposed method performs well in finite samples.
Read the full abstract
Geometric Topology
A central limit theorem for the signatures of 2-bridge knots
Cohen, Lowrance, Madras, and Raanes computed the average (absolute value of) signature over all 2-bridge knots with crossing number $c$ by introducing the number $s(c,\sigma)$ of 2-bridge knots of crossing number $c$ and signature $\sigma$. Here we provide a closed formula for this number. We use these calculations to show that the distribution of the signatures of 2-bridge knots with crossing number $c$ approaches a normal distribution as $c$ tends to infinity.
Read the full abstract
Numerical Analysis · Cross-listed from cs.LG
A Hybridizable Neural Time Integrator for Stable Autoregressive Forecasting
For autoregressive modeling of chaotic dynamical systems over long time horizons, the stability of both training and inference is a major challenge in building scientific foundation models. We present a hybrid technique in which an autoregressive transformer is embedded within a novel shooting-based mixed finite element scheme, exposing topological structure that enables provable stability. For forward problems, we prove preservation of discrete energies, while for training we prove uniform bounds on gradients, provably avoiding the exploding gradient problem. Combined with a vision transformer, this yields latent tokens admitting structure-preserving dynamics. We outperform modern foundation models with a $65\times$ reduction in model parameters and long-horizon forecasting of chaotic systems. A "mini-foundation" model of a fusion component shows that 12 simulations suffice to train a real-time surrogate, achieving a $9{,}000\times$ speedup over particle-in-cell simulation.
Read the full abstract
Mathematical Physics · Cross-listed from quant-ph
A rigorous quasipolynomial-time classical algorithm for SYK thermal expectations
Estimating local observables in Gibbs states is a central problem in quantum simulation. While this task is BQP-complete at asymptotically low temperatures, the possibility of quantum advantage at constant temperature remains open. The Sachdev-Ye-Kitaev (SYK) model is a natural candidate: at any constant temperature, its Gibbs states have polynomial quantum circuit complexity and are not described by Gaussian states. Rigorous analyses of the SYK model are difficult due to the failure of known techniques using random matrix theory, cluster expansions, and rigorous formulations of the quantum path integral and replica trick. Despite this, we give a rigorous proof of a quasipolynomial-time classical algorithm that estimates SYK local thermal expectations at sufficiently high constant temperature. Our result introduces a new Wick-pair cluster expansion that we expect to be broadly useful for disordered quantum many-body systems.
Read the full abstract
Mathematical Physics
A Nearest-Neighbor Hard-Core Model on a Penrose Graph
We prove that the maximal graph-density of an independent set in a Penrose P3 tiling considered as a planar non-directed graph is equal to $(57 - 25 \sqrt{5})/2 \approx 0.54915$ despite the fact that the graph is bipartite. Accordingly, the extreme Gibbs measure of the nearest-neighbor hard core particle model on this graph is unique for sufficiently large values of the particle activity. This invalidates a natural expectation to observe the coexistence of even and odd phases.
Read the full abstract
Numerical Analysis
Old and new Schrödinger eigenvalue localisation
Unconditional guaranteed lower and upper eigenvalue bounds are mandatory for the understanding of the Schrödinger eigenvalue spectrum and its spectral gaps. While upper eigenvalue bounds are naturally induced by conforming discretisations, guaranteed lower eigenvalue bounds (GLB) are less immediate. This paper clarifies the adaptation of nonconforming GLB from the harmonic eigenvalue problem and discusses their comparison for general and piecewise constant potentials. A fine-tuned extra-stabilised scheme is proposed and found superior in numerical comparisons. This new direct calculation of GLB is compatible with adaptive mesh-refinement and successfully circumvents the appearance of maximal mesh-size parameters in former GLB based on post-processing. Computational benchmarks also investigate guaranteed upper eigenvalue bounds (GUB) for two-sided eigenvalue control by conforming test functions associated to the underlying nonconforming computations. A numerical comparison with GUB from additional lowest-order conforming finite element schemes shows competitive accuracy with less computational cost.
Read the full abstract
Dynamical Systems · Cross-listed from eess.SY
On the dynamic behavior of the network SIRS epidemic model
We study the Suscectible-Infected-Recovered-Susceptible (SIRS) epidemic model on deterministic networks. For connected but otherwise general interaction patterns and heterogeneous recovery and loss-of-immunity rates, we identify a fundamental parameter R_0 (the basic reproduction number), which fully characterizes the qualitative dynamic behavior of the system. This parameter is the dominant eigenvalue of a rescaled version of the interaction matrix, whose rows are normalized by the corresponding recovery rates. We prove that a transcritical bifurcation occurs as R_0 crosses the threshold value 1. Specifically, we show that, if R_0 does not exceed 1, then the disease-free equilibrium is globally asymptotically stable, whereas, if R_0 is larger than 1, then the disease-free equilibrium is unstable and there exists a unique endemic equilibrium, which is asymptotically stable. As a byproduct of our analysis, we also identify key monotonicity properties of the dependence of the endemic equilibrium on the model parameters (the interaction matrix as well as the recovery rates and the loss-of-immunity rates) and obtain a distributed iterative algorithm for its computation, with provable convergence guarantees. Our results extend existing ones available in the literature for network SIRS epidemic models with rank-one interaction matrices and homogeneous recovery rates (including the single homogeneous population SIRS epidemic model).
Read the full abstract
Optimization and Control
Modeling Gaps in Hydropower Cascading System Models: A Systematic Review of Rule-Based Formulations
The coordination of cascading hydropower systems represents a fundamental challenge in modern energy systems engineering, requiring a sophisticated balance between multi-reservoir physics, stringent environmental regulations, and dynamic market participation. As intermittent energy sources increase, the transition to high-fidelity hydropower modeling has become a core requirement for ensuring power system reliability, long-term energy resilience and affordability. This review provides a comprehensive analysis of 131 seminal articles through the exclusive lens of optimization-based approaches, intentionally omitting pure simulation and heuristic methods to focus on rigorous mathematical formulations. A generalized 10-equation mathematical framework is established as a formal baseline, capturing the full physical and hydraulic behavior of cascading systems, including spatiotemporal inflow routing, storage-to-elevation relationships, head-dependent power generation, and prohibited operating zones. Each article is evaluated against every equation of this Standard Model in a systematic census documenting which physical relationships are included, simplified, or omitted, providing an empirical measure of modeling fidelity across the field. Modeling simplifications are evaluated through the lens of grid reliability rather than water management performance alone. The review makes a focused technical case for mixed-integer linear programming with piecewise linear approximations as the optimal balance between physical accuracy and computational tractability, highlighting recent optimal regression techniques that minimize combinatorial overhead. Finally, a bibliometric analysis of solver usage identifies the near-absence of high-performance open-source solvers as a critical reproducibility barrier, and a promising avenue for broader adoption of high-fidelity cascading hydropower models.
Read the full abstract
Geometric Topology
Uniform spectral gap of scl in $2$-orbifolds
We show a uniform spectral gap of stable commutator length for all compact hyperbolic $2$-orbifolds relative to the peripheral subgroups. Except for the case of a sphere with three cone points, we have an explicit uniform gap $1/36$. These estimates are needed in understanding stable commutator length in $3$-manifolds. Our methods use explicit quasimorphisms for the generic case, and use hyperbolic geometry (pleated surfaces) for the exceptional case of a sphere with three cone points.
Read the full abstract
Numerical Analysis
Data-Driven Surrogate Models for Agromaritime Applications: Finite Element-Neural Network Integration
Predicting nutrient transport and salinity distribution is crucial for mitigating climate-related threats to agromaritime systems. Traditional PDE-based models can capture the physics of nutrient dispersion, salinity and water quality. However, they face challenges in scalability and adaptability to real-time problems. In this article, we develop a hybrid approach that combines finite element discretisations with neural network integration to enable efficient and adaptive data-informed predictions. We use a finite element solver for the steady-state diffusion-reaction equation to generate a dataset across varying diffusivity, reaction and inflow conditions. We then build a proper orthogonal decomposition (POD), which reduces dimensionality, and a neural network (NN) that maps parameters to reduced coefficients. A numerical study presented on a simplified model demonstrates the proof-of-concept for predicting nutrient transport and salinity distribution. Numerical experiments show that the NN surrogate achieve a speed-up of approximately 956x compared to a regular FEM solver while maintaining an accuracy of mean relative L2-errors of 15% across the test set, with occasional higher deviations, which is sufficient for rapid scenario screening and parametric studies. These results highlight the method's potential as a fast and accurate surrogate for nutrient and salinity prediction, offering a balance between FEM reliability and NN adaptability for sustainable agromaritime management.
Read the full abstract
Functional Analysis
A Bochner-type integration theory for random normed modules
We develop a measure and integration theory for random normed modules. Given a probability space $({\rm X},\Sigma,\mathfrak m)$, we introduce and study measures taking values into the space $L^0(\mathfrak m)$ of $\mathfrak m$-measurable functions quotiented up to $\mathfrak m$-a.e. equality. Moreover, we develop a Bochner-type integration theory with respect to an $L^0(\mathfrak m)$-valued measure $\mu$, for maps whose target ${\rm M}$ is a complete random normed module with base $({\rm X},\Sigma,\mathfrak m)$, or equivalently an $L^0(\mathfrak m)$-Banach $L^0(\mathfrak m)$-module. Inter alia, we prove versions of the Radon-Nikodým theorem and of the Riesz-Markov-Kakutani representation theorem for $L^0(\mathfrak m)$-valued measures. We also outline several applications of our integration theory: we introduce a notion of martingale with values in a complete random normed module, we propose a definition of random Radon-Nikodým property and we discuss random sets of finite perimeter.
Read the full abstract
Dynamical Systems
Emergence of Mandelbrot-like and Julia-like Structures in Parameter Slices of Rational Maps
We study complex one-dimensional parameter slices in a three-parameter family of rational maps with two free critical points, obtained by imposing the existence of periodic orbits with prescribed multipliers. Using explicit parametrizations, we explore these slices numerically by analyzing the behavior of the critical orbits and approximating the corresponding connectedness loci. The computations reveal rich parameter space structures closely analogous to those arising in cubic polynomial families, including Mandelbrot-like sets. In addition, we observe regions exhibiting Julia-like structures embedded in parameter space, arising from the interaction between bounded and escaping critical orbits. While the appearance of such structures is well established in polynomial dynamics, it remains comparatively less explored in the setting of rational maps. Our results provide numerical evidence that these parameter slices contain subsets closely related to the period-one and period-two slices of cubic polynomial families. More precisely, certain regions appear to exhibit geometric and dynamical features consistent with embedded copies of these classical parameter spaces. These observations highlight how classical structures from polynomial dynamics can emerge naturally within parameter slices of rational maps.
Read the full abstract
Analysis of PDEs
Dimension Drop for Harmonic Measure on Ahlfors Regular Boundaries
We provide quantitative estimates for the dimension drop of harmonic measure. We show that for a domain $\Omega = \mathbb{R}^{n+1} \setminus E$ where $E$ is an $s$-Ahlfors regular compact set satisfying a uniform $L^2$-based non-flatness condition $\beta_2 \ge \delta_0$, the dimension of its harmonic measure is strictly less than $s$ for $s \in (n - c\delta_0^2, n]$. For planar domains, we establish an analogous quantitative threshold $s_0 = 1 - c\delta_0^2$ under Azzam's uniform non-flatness condition $\beta_\infty + \beta_{\operatorname{hole}} \ge \delta_0$.